Introduction:

Background/Objectives: ¹⁷⁷Lu-PSMA is a novel targeted radioligand therapy (RLT) that is FDA-approved for treatment of metastatic castration-resistant prostate cancer (mCRPC). Treatment eligibility is based on positive ⁶⁸Ga-PSMA-PET imaging but this cannot predict who will respond. This study investigated whether changes in PROSTest, a blood-based molecular assay utilizing prostate cancer-specific transcripts, after one cycle of therapy could predict treatment response to continued therapy by ¹⁷⁷Lu-PSMA in mCRPC patients. We hypothesized that a decrease in PROSTest score after first cycle of treatment from baseline would be associated with response (PFS – PSA nadir) and overall survival (OS). Our goal was to determine who should continue with therapy.

Methods:

We evaluated results from two hundred and sixty-eight patients (median age: 73 years, range: 58-89) with mCRPC from 3 different centers (Munster, Basel and Essen) were treated with a median of 2 (range 1-6) cycles of ¹⁷⁷Lu-PSMA as qualified by PET. PROSTest scores were measured at baseline and after one treatment cycle. Samples for PSA were collected prior to every cycle and PSA measured using standard clinical assays. Response (PFS) was categorized by PSA nadir (> -50% decreases). Changes in PROSTest (baseline vs. after 1 cycle) were correlated with PFS or OS using AUROCs, Kaplan-Meier survival analysis and hazard ratios.

Results:

There were 133 responders and 135 non-responders (based on PSA nadir); 130 patients died from their disease. Median PFS was 15.0 months [95% CI: 8.7–15.0] in responders vs. 4.4 months [3.4–6.1] in non-responders. OS was 8.6 months [7.4–11.2] in deceased patients (n=130) and undetermined in survivors (n=135). Kaplan-Meier analyses show that changes in PROSTest were predictive for assessing both PFS and OS. For PFS, patients with a decrease in PROSTest score had a markedly better response (mPFS 15.0 months [8.7-15]) compared to those with increased scores (mPFS: 4.4 months [3.4-6.1]). PROSTest demonstrated a strong association with PFS (log-rank Chi²=31.4, p<0.0001; HR=0.37; [0.36–0.71]). PROSTest also demonstrated a strong association with OS (log-rank Chi²=56.4, p<0.0001; HR=0.26 [0.18–0.37]). Notably, 90.4% (122/135) of patients who clinically responded with a demonstrable survival advantage (mOS undefined) showed a decrease in PROSTest, while 97 of 130 (74.6%) patients who progressed had an increase (mOS: 8.6 months [8.8-14.3]), underscoring the test’s utility as a sensitive, early predictor of long-term therapeutic benefit.

Conclusion:

This audit identifies that early changes in PROSTest are significantly associated with both PFS and OS for mCRPC undergoing ¹⁷⁷LuPSMA RLT. Importantly, PROSTest often provided prediction insight earlier than conventional imaging or PSA kinetics. These multi-center findings identify PROSTest as a minimally invasive blood-based assay for early prediction of response to LuPSMA therapy, and inform treatment personalization in mCRPC.

Funding: N/A